Inter-turn faults in synchronous machines can lead to catastrophic failure of the machine. The time elapsed in the growth of an incipient fault leading to a catastrophic failure of the machine varies from a third of a second to several minutes in the case of random-wound machines and can even be much longer in the case of form-wound machines. Hence, early detection of such faults involving few turns is desirable in order to protect them.
U.S. Pat. No. 6,611,771 discloses a stator turn fault detection system and method capable of real-time detection of a stator turn fault in an electric motor. The stator turn fault detector includes a feed forward neural network that when trained, using fundamental frequency sequence components of the voltage and current supplying the electric motor, will estimate a fundamental frequency sequence component of current indicative of a stator turn fault. A method for detecting a stator turn fault in an electric motor as well as a method for training a feed forward neural network for use with the stator turn fault detector is also disclosed.
The use of harmonics to detect inter-turn faults has been disclosed in Penman and Jiang, “The detection of stator and rotor winding short circuits in synchronous generators by analyzing excitation current harmonics,” International Conference on Opportunities and Advances in International Electric Power Generation, Conf. Publ. No. 419, pp. 137-142, March 1996. However, some of the even harmonics described in this paper in the field current for detection of stator fault can also increase with supply unbalance, and can lead to serious confusion in synchronous machine monitoring. These harmonics can be induced in the two-coil or four coil (standard) field windings. Therefore, not all harmonics can be used reliably to detect inter-turn faults. Accordingly, there is a need for a method to determine the harmonic or harmonics that can be used to detect inter-turn faults.
It is an object of the present technology to overcome the deficiencies in the prior art.